This is a bad analogy. You don't need to copy title and chapter names to allow other people to "interoperate" (read, scan, and index I guess?) with the book. Moreover substantial creativity goes into title and chapter names. Both of which imply they should be substantially more copyrightable than APIs.
> This is a bad analogy. You don't need to copy title and chapter names to allow other people to "interoperate" (read, scan, and index I guess?) with the book.
Bad analogies are inevitable, and aren't the commenter's fault. It's the Federal Circuit's fault, for trying to blur the lines between functional matters (the domain of patents) and copyright matters. Any analogy that's simple enough to quickly understand will suffer from basically this same flaw, or else not apply to this case.
Oracle's trying to get copyright duration on the exclusive rights to something that falls under patent subject matter, so yeah, they're blurring the lines.
Ultimately, the problem for Oracle is that the APIs are too functional to be eligible for copyright protection, too abstract to be eligible for patent protection, and too generic to be eligible for trademark protection. But that doesn't stop them from trying to get the best features from all the above.
